Aux/Line-in recording issue?

[BJ]

I had a JB7 for a good few years and have now replaced it with a B2 as it was running out of room.  My days of line-in recording (of LPs, etc.) are mostly past but I sometimes record live concerts (classical) from digital radio to an HD AV recorder.  This and the connected HD TV are plumbed into my old analog stereo system via the headphone output on  the TV.  Playing back the recording I could via the amp record the music (in real time) to the JB7.  This produced files that sound good to me (and stand comparison to ripped CDs).  I use Audacity to trim the recordings and break them into tracks.  Audacity displays them in a waveform.

The waveforms of the few recordings I have made on the B2 seem to my far from expert eye all to have a certain flattening of negative signals (below the 0.00 line) on both channels.  I've no idea what the waveform is measuring exactly but my impression is that the JB7 MP3 files displayed (on average) roughly equal + or - peaks.  

For testing, I have taken an optical feed straight from the HD recorder and run this into the B2 through a basic DAC and again the results are the same, so presumably it's not the (no longer young) AV or HiFi kit failing.  (Is the B2 likely to get an optical audio input any tme soon?)

I have also copied recordings from the HD recorder to DVD which allows PC access to the AV files, from which Audacity can extract the audio tracks.  I have edited these and compressed them (FLAC) — their waveforms do not seem to show the same negative flattening.  I've looked at various instrumental combinations (orchestral, through chamber groups to solo) and the waveform asymmetry  seems to be common to all.  I'm attaching images which seem to show (though the amplitude varies, which doesn't help!) not so many negative peaks at the bottom of the wave:  the B2 line-in recordings are the bottom pair.

It's difficult to set up parallel listening tests so I'm not sure how much difference I can actually hear and it may all just be suggestibility.  Nor do I  know quite how the B2's Aux recording system works but does the waveform suggest something isn't quite right?  

Thanks for reading all this!

BJ

[Daniel Taylor]

BJ,

I looked only at your first picture.  For discussion, consider the section beginning at about the 7:45 time through about 9:30.  I do see the somewhat reduced peaks of the B2 on the negative going part of the bottom two traces.  But I also see what are noticably higher peaks on the positive going part of the B2 traces.  This seems very strange indeed.  I don't even know the correct words to describe the phenomenon.  Nor can I suggest in what way it would be audible.  But it is certainly not right.  I'd expect that an audio electronics engineer could tell us what is going wrong in the circuitry.

[Daniel Taylor]

So I went on and looked at the second picture.  Looking at the burst just past the 1:50 mark.  To me it looks like the channels for the lower pair of traces are reversed from the upper pair.  That is to say that the very top trace represents the same audio channel as the very bottom trace.  No big deal, but pertinent when comparing the wave shapes.

Also, looking closely at the burst, the very bottom trace seems to be inverted polarity to the corresponding very top trace.  (No way to tell which is correct.)  The top trace seems to be ever so slightly more negative going on the leading edge, but is mostly symetrical about the zero axis.  But the very bottom trace seems to be more positive going and noticably less symetrical about the axis.

You may have done this already, but to know whether the fault is with the B2's input circuit or with the output circuit of your HD recorder, you'd have to input something to the B2 from a different source and see if you get the same symptoms.

[BJ]

Thank you — reassuring that I may not be dreaming the 'flattening' effect, even if it remains perplexing.  And it needed another pair of eyes to see the 'inversion' effects.

I don't think I have a way of having the B2 record from a different source direct (don't have much kit left these days) but I fed a CD player output into the B2 via my amp.  The result was weird and I discounted it as an aberration.  But I've now done it again a few times and it still looks weird.  Couldn't get the recording level any lower to find the top of the extended + spikes.  And the - side never gets below -0.5...

The image compares it with the B2-ripped CD.

The thoughts of an (ideally 'official') "audio electronics engineer" would indeed be most welcome...

BJ

[Daniel Taylor]

BJ,

If it hurts when you do that, then don't do that.  ;-)

I do all may recording from analog sources on my computer.  Then I can split tracks, maximize levels, and sometimes add a tiny bit of compression using Audacity (excellent free software).  I would not bother with the B2's analog recording ability.  But it's a possibility for those who have no other option.  Since you already have Audacity, I'd think this would be the likely thing to do.

[BJ]

Assuredly, extracting the audio from the A/V recordings on the HD recorder and processing it through the invaluable Audacity is the way I shall go for now.  However, line-in recording is a method offered by the B2 and Brennan may wish to know that it seems capable of misbehaviour.  And they may be able to establish quite what that misbehaviour is.  Also, with my setup at least,  line-in has the advantage that I can hear the recording as it goes in and note where editing will be necessary.  Copying the HD recorder files to DVD is an in-machine process (but still in real time) which necessitates either a subsequent listen-through or blind editing on the fly.

I really would appreciate some Brennan comment on this.

BJ

[Daniel Taylor]

Well said.  I hope Brennan takes an interest.

[PMB]

Hi BJ,

I have spoken to Martin and he thinks it could be due to signal levels i.e. causing the B2's ADC to clip.

Have you been able to hear a difference between the two sources?

I have replied further to your email.

Paul

Brennan Support